We have conducted several phase I studies to evaluate the antigenicity and reactogenicity of various formulations of the rhesus rotavirus-based quadrivalent vaccine candidate in two month old infants. In these studies we aimed at achieving an arbitrary 50% "take rate" standard for each rotavirus serotype in the vaccine. In general, increase in the vaccine titer and the number of doses resulted in greater antigenicity. The quadrivalent vaccine was associated with mild, short-lived febrile episodes in about 25% of the infants tested. This level of reactogenicity was considered acceptable. A three dose schedule of the quadrivalent vaccine at a titer of 1 x 105 PFU of each component best approached the desired standard and is being evaluated in an efficacy trial in Venezuela. We have also investigated the M37 rotavirus as an alternative vaccine candidate alone or in combination with RRV-based quadrivalent vaccine. M37 was safe but less immunogenic than the quadrivalent vaccine.